Received 14 March 1994/Returned for modification 31 May 1994/Accepted 1 August 1994 ... cultured and analyzed daily for gram-positive and gram-negative bacteria. ..... Unertl, K., G. Ruckdeschel, H. K. Selbman, U. Jensen, H. Forst,.
ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Nov. 1994, p. 2564-2567
Vol. 38, No. 11
0066-4804/94/$04.00+0
Copyright ©D 1994, American Society for Microbiology
Bacteria:
a
Major Pathogenic Factor for Anastomotic Insufficiency
H. M. SCHARDEY,l* T. KAMPS,1 H. G. RAU,' S. GATERMANN,2 G. BARE1TON,3 AND F. W. SCHILDBERG'
Department of Surgery, University Hospital Grosshadem,r and Department of Pathology, 3 Ludwig-MaximiliansUniversity, Munich, and Department of Microbiology, University of Lubeck, Libeck 2 Germany Received 14 March 1994/Returned for modification 31 May 1994/Accepted 1 August 1994
The aim of this study was to determine the influence of bacteria on the development of anastomotic insufficiency following gastrectomy in the rat. Fifty-seven male Wistar rats were randomly assigned to three groups and subjected to gastrectomy. Group I (n = 20) was orally inoculated with 109 Pseudomonas aeroginosa organisms on postoperative day 1. Group II (n = 20) served as the control group. Group III (n = 17) was decontaminated with 320 mg of tobramycin, 400 mg of polymyxin B, and 500 mg of vancomycin per liter of fluid administered from preoperative day 7 to postoperative day 10. Swabs from the oropharynx and rectum were cultured and analyzed daily for gram-positive and gram-negative bacteria. Surviving animals were sacrificed on postoperative day 10. All animals were autopsied immediately following death. Anastomotic insufficiency was defined as a histologically proven transmural defect at the suture line. Along with an effective reduction of pathogenic bacteria colonizing the oropharynx, the rate of anastomotic insufficiency could be reduced significantly, to 6% in decontaminated animals compared with 80% in controls (P < 0.001 by Fisher's exact test). Inoculation of group I animals with P. aeruginosa led to an increase of anastomotic insufficiency up to 95% and a significant increase in mortality (P < 0.05). We conclude that bacteria play a major role in the pathogenesis of anastomotic insufficiency following gastrectomy in the rat.
of polymyxin B, and 500 mg of vancomycin per liter of fluid (administered orally). All rats were operated on under sterile conditions with intraperitoneal pentobarbital anesthesia (50 mg/kg of body weight). The stomach was completely removed. The continuity of the intestinal tract was restored by an end-to-end esophagoduodenostomy. Anastomoses were performed by employing 6X0 PDS and a transmural, single-layer running suture technique. On postoperative day 10, surviving animals were sacrificed. All animals were autopsied. The anastomoses and the lungs were histopathologically examined. The mechanical resistance of the anastomosis was determined by measuring the bursting pressure in situ. NaCl (0.9%) in a 50-ml syringe was pumped into the system at 2 ml/min with a perfusor pump (Braun; maximum producible pressure, 300 mm Hg [ca. 40,000 Pa]). The pressure changes were detected (Stadham ID23 pressure transducer) and recorded (Hellige monitor and writer). Swabs for bacteriological analysis were collected daily from the oropharynx and rectum. Samples were immediately plated and incubated on MacConkey agar and blood agar and analyzed according to the guidelines of the German Society for Hygiene and Microbiology (4). Results were recorded as no growth or as low, moderate, or massive bacterial growth. Tobramycin and vancomycin in the urine and serum samples from group III animals were analyzed by a fluorescence polarization immunoassay (11). The drug resistance of bacteria isolated from group III animals was determined by agar diffusion according to Deutsche Industrie Norm (4). The German National Research Council's guide for care and use of laboratory animals was followed. Definitions. Anastomotic insufficiency was defined as a histologically proven transmural defect at the anastomotic suture line with a vital inflammatory reaction, e.g., exudation of fibrin and leukocytes. Statistical methods. Means and standard deviations were computed, and the Mann-Whitney test, the Kruskall-Wallis test, Fisher's exact test, and the chi-square test were used. A value of P < 0.05 was considered to be significant.
Anastomotic insufficiency is one of the major causes of morbidity and mortality following total gastrectomy. The pathogenesis of esophago-intestinal anastomotic insufficiency, however, is not completely understood. Impaired blood supply or local microcirculatory disturbances leading to necrosis (13), as well as foreign bodies (6, 7), have been thought to be the cause of anastomotic insufficiency for more than 100 years. Anastomotic insufficiency, however, is a septic disease. Exogenous or endogenous potentially pathogenic microorganisms colonizing the digestive tract may play a causative role in the pathogenesis in addition to microcirculatory disturbances. This idea led us to examine the influence of bacterial colonization on the incidence of anastomotic insufficiency. By topical application of nonresorbable bactericidal antibiotics (16, 18), the colonizing microflora of the oropharynx and upper gastrointestinal tract can be easily manipulated. We therefore used deliberate colonization and topical decontamination with tobramycin, polymyxin and vancomycin to test the following hypotheses in an experimental study: (i) anastomotic insufficiency occurs as a result of bacterial infection, and (ii) anastomotic insufficiency can be prevented by preventing bacterial colonization.
mg
MATERUILS AND METHODS Fifty-seven male Wistar rats were randomly assigned to three groups. In group I (bacterial inoculation), all animals received one oral dose of 109 Pseudomonas aeruginosa organisms on the first postoperative day. P. aeruginosa was chosen because it is commonly causes nosocomial infections in humans (5, 8, 9). Group II animals served as the controls. Group III animals were decontaminated from preoperative day 7 to postoperative day 10 by addition of 320 mg of tobramycin, 400 * Corresponding author. Mailing address: Department of Surgery, University Hospital Grosshadern, Ludwig-Maximilians-University, Marchioninistr. 15, D-81377 Munich, Germany. Phone: 00-49-89-70952423. Fax: 00-49-89-7095-8893.
2564
BACTERIA AND PATHOGENESIS OF ANASTOMOTIC INSUFFICIENCY
VOL. 38, 1994
TABLE 2. Semiquantitative analysis % of samplesa:
TABLE 1. Qualitative analysis % of samples positive
SpeciesOcultures Group Group Group Group Group III II I (n = 130)' (n = 165) (n = 166)
Oral Gram negative
I,
Escherichia coli Proteus sp. Pasteurella multocida Kiebsiella sp. P. aeruginosa Flavobacterium odoratum Streptococcus sp. Staphylococcus sp. S. aureus Corynebacterium sp. a
33.8 33.0 23.8 3.0 37.6 74.8 13.0 0.8 0.8
56.3 29.6 39.6 4.2
0.6 87.8 53.9 1.3
Group
Culture and strains
_
+
++
+++
I (n = 130)b II (n = 165) III (n = 166) I II III
27.3 20.6 71.6 16.1 7.7 77.1
31.2 42.4 20.4 32.2 16.9 17.4
31.2 21.8 4.8 36.9 29.0 3.6
10.5 14.5 2.4 13.1 45.4 1.2
II III II III
0.6 58.7 93.1 66.2
13.7 0.7 29.6
8.3 14.8 4.8 8.7
91.0 12.8 1.3 5.4
Rectal
Oral cultures
9.0 16.3 4.8 0.6
13.3 5.4 4.2 0.6
2565
95.1 53.0
4.7 35.8
Gram positive
0.6
1.3
Rectal Gram negative
7.5 0.6
29.0 8.1 0.6 0.6
n, number of samples.
RESULTS
Colonization rates. During the postoperative period, dynamic changes in the bacterial colonization pattern in the oropharynx were observed for all groups. In groups I and II there was an increase in gram-negative organisms and a transient decrease in gram-positive organisms. At the time of operation and during the entire observation period, the colonization rates in decontaminated animals stayed well below those in groups I and II. Species of bacteria isolated. The spectra of bacteria species isolated from the oral cavity were similar for all groups (Table 1), with the exception of P. aeruginosa, which was artificially introduced into group I. Even though P. aeruginosa could be isolated from 95% of the animals in group I at one time or another during the monitoring, only 37.6% of oral samples were positive for P. aeruginosa. Decontamination caused a change in the composition of the oral microflora along with a marked reduction of most bacterial strains. Mixed colonization with gram-negative and gram-positive bacteria was observed with similar frequencies in groups I (71%) and II (75%) but was a rare event in group III (11%); 12.1% of group I, 18.5% of group II, and 2.7% of group III animals showed the presence of more than one gram-negative strain or more than one gram-positive strain. Semiquantitative analysis. In groups I and II the percentages of oral swabs yielding moderate or massive growth of gram-positive and gram-negative bacteria were higher than that in the decontaminated group (Table 2). Influence of decontamination on rectal flora. Rectal colonization patterns also differed in the decontaminated and control groups, showing both semiquantitative and qualitative reductions of gram-negative bacteria and a shift towards grampositive species (Tables 1 and 2). Drug resistance. The development of drug resistance after 18 days of decontamination was not observed. Bursting pressure. The bursting pressure of anastomoses in decontaminated animals (group III) was highest, at 247 81) (mnean + standard deviation) mm Hg (ca. 32,900 + 10,800 Pa). The bursting pressure in control animals (group II) was 178 141 mm Hg (ca. 23,700 ± 18,800 Pa), and that in group I animals was 99 + 97 mm Hg (ca. 13,200 12,900 Pa). The bursting pressures in groups II and III did not differ significantly (P = 0.478; Mann-Whitney test), while the values for
Gram positive
a-, no bacterial growth; +, low bacterial growth; + +, moderate bacterial growth; + + +, massive bacterial growth. b n, number of samples.
group I animals were significantly lower than those for both other groups (P = 0.0075; Kruskall-Wallis test). Anastomotic insufficiency. Eighty percent of the specimens from the control group had an anastomotic insufficiency on histologic examination, whereas this was seen with only one animal of group III (6%). This reduction in the rate of anastomotic insufficiency was statistically highly significant (P < 0.001 by Fisher's exact test). Inoculation with P. aeruginosa in group I led to an anastomotic insufficiency rate of 95%, with only one animal left with an intact anastomosis (Table 3). The only insufficiency in the decontaminated group occurred in an animal with a small transmural abscess. Free perforation with peritonitis was not observed. In the control group (group II), insufficiencies with peritonitis occurred in two animals (10%); both died by day 2. The other 13 animals with insufficiencies had transmural abscesses covered by granulation tissue with the beginning of a scarring reaction. In the P. aeruginosa-contaminated group (group I), five animals (25%) died from insufficiency with free perforation and peritonitis, and one died from secondary rupture of a very large transmural abscess followed by peritonitis on day 8. In all cases of peritonitis, P. aeruginosa could be recovered. One animal had a large transmural abscess covered by granulation tissue with the beginning of a scarring reaction, and the 12 remaining animals (60%) with anastomotic insufficiencies had extensive fistulas into the liver. On histologic examination, all insufficient anastomoses from all groups presented with pus. Gram staining of the specimens demonstrated the presence of bacteria on all but two anastomoses. Both were from group II animals; one was insufficient TABLE 3. Complications % in: Complication
Group I (n = 20)
Group II (n = 20)
95 30 10 80 30 30
80 10 30 50 15 10
Insufficiency Peritonitis Abscess Adhesions Pneumonia Mortality a Significant difference (P
